Research commentary

Rehab that follows you home the HEAD pilot.

Most rehabilitation ends when the prescription runs out, and for chronic stroke survivors what follows is often slow decline. An Italian multicenter team asked whether tele-monitored VR could carry rehabilitation past the clinic's front door, and tracked what happened for six months after. The most interesting number in the study isn't a gain. It's what didn't decay.

Reviewed by The Karuna Labs clinical teamUpdated

Can VR rehabilitation work at home for people with chronic stroke?

The 2021 HEAD pilot suggests it can be delivered there. Thirty-four chronic stroke survivors improved across walking, balance, and cognition during a supervised clinic VR phase; eleven then continued with tele-monitored home VR for three months, completing 89% of sessions with no adverse events. At six months, the home group held onto functional mobility while the usual-care group declined significantly.

It is a small pilot with an uneven comparison group, so the durability finding is a signal to test, not a proven effect. But feasibility, the question pilots exist to answer, came back emphatically yes.

At a glance

Article analysed
Jonsdottir et al. (HEAD Study Group), Frontiers in Neurology, 2021; registered as NCT03025126
Who was studied
34 outpatients at least 6 months post-stroke (mean age 59), across multiple Italian centers
Design
Two phases: uncontrolled clinic VR phase for all 34, then 1:2 allocation to home VR (n = 11) or usual care (n = 23) for 3 months, single-blind assessment
Clinic-phase results
Significant gains in 2-minute walk (p = 0.011), 10-meter walk (p = 0.006), Berg balance (p = 0.010), MoCA cognition (p = 0.001), memory (p = 0.031), with large effect sizes
Home-phase result
No significant between-group differences at 3 months; at 6 months the home group maintained walking capacity (p = 0.032) while usual care declined (p = 0.02)
Adherence
92% of clinic sessions; 89% of home sessions (~55 of 60 planned); mean 35.3 minutes of VR daily
Safety and usability
No study-related adverse events; System Usability Scale median 77.5/100

Key takeaways

  • The standout result is adherence: 89% of prescribed home sessions completed over three months, averaging 35 minutes daily. Home exercise programs on paper handouts rarely see numbers like that.
  • The clinic-phase improvements came from an uncontrolled before-after design in chronic stroke; without a control group they are encouraging, not attributable.
  • The six-month pattern, home group holding walking capacity while usual care declined, points at maintenance as home VR's natural role: not accelerating recovery, but stopping the slide after formal rehab ends.
  • Eleven versus twenty-three, allocated by equipment availability rather than balanced randomization, with a passive control: every efficacy signal here needs a properly powered trial behind it.
  • The delivery model, headset at home, remote monitoring, clinician oversight, is the same shape as modern virtual chronic pain care, and this study is among the better feasibility evidence that the model holds together.

What did the HEAD study actually do?

The HEAD program (Human Empowerment Aging and Disability) built a VR system for combined motor and cognitive rehabilitation, and tested it in two stages. First, all 34 participants, chronic stroke survivors at least six months out, past the window of spontaneous recovery, completed a supervised VR program in the clinic. Then eleven took the system home for three months of tele-monitored use, while twenty-three continued with usual care, and everyone was assessed out to six months by blinded raters.

The uneven 1:2 split was pragmatic, allocation was limited by available equipment, and it is one of the study's acknowledged weaknesses. The daily home practice mixed exercise types: on average 35.3 minutes of VR per day, split across motor (18.6 minutes), cognitive (10.7), and occupational (6.3) tasks.

Assessment covered walking capacity and speed (2-minute and 10-meter walk tests), balance (Berg Balance Scale), cognition (Montreal Cognitive Assessment), and memory (Rivermead Behavioural Memory Test), a broader net than most motor-rehab trials cast.

What did the results show, phase by phase?

Clinic phase: broad gains, no control

After the supervised clinic program, the 34 participants improved significantly on nearly everything measured: 2-minute walk distance (p = 0.011, effect size d = 0.894), 10-meter walk speed (p = 0.006, d = 1.071), Berg balance (p = 0.010, d = 1.067), MoCA cognition (p = 0.001, d = 1.253), and memory (p = 0.031, d = 0.817). In chronic stroke, where the natural trajectory is flat or declining, gains of this breadth are notable, but with no control group, expectation, attention, and extra activity of any kind remain unruled-out explanations.

Home phase: no separation at three months, then the follow-up story

During the three-month home intervention itself, change scores did not differ significantly between the home-VR and usual-care groups. The interesting divergence appeared at six months: the home group maintained its functional mobility (2-minute walk, p = 0.032 versus usual care), while the usual-care group's walking distance declined significantly (p = 0.02). The home group also improved on memory (p = 0.001).

Read as a trajectory rather than a snapshot: everyone gained in the clinic; those sent home with nothing began giving the gains back; those sent home with the system kept them. That is exactly the pattern you would predict if the home VR's job was maintenance dose.

Why is the adherence finding the important one?

Because adherence is where home rehabilitation dies. The standard of care after discharge is a home exercise sheet, and the quiet consensus in rehabilitation is that most of those sheets go un-exercised. Against that baseline, 89% completion of sixty prescribed home sessions, roughly 55 sessions actually done, with 35 minutes of daily practice, is a striking behavioral result in its own right.

The supporting numbers cohere: 92% adherence in the clinic phase, a median System Usability Scale score of 77.5 out of 100 from a cohort with a mean age of 59 using a technology stereotyped as belonging to their grandchildren, and zero study-related adverse events across both phases.

This matters beyond stroke because dose and consistency are the currency of every retraining-based therapy. Whatever the eventual verdict on VR's specific efficacy, a delivery channel that multiplies completed practice is valuable, the same argument that surfaced in the engagement data of the subacute stroke meta-analysis.

What does this mean for remote rehabilitation and chronic pain care?

The study's limitations should be listed without flinching: a small, equipment-constrained allocation; a passive control that received no comparable attention; eligibility that excluded significant cognitive impairment and required standing tolerance; and follow-up capped at six months. Every efficacy signal here is a hypothesis for the larger trial the authors themselves call for.

What the pilot establishes more firmly is the model: clinic onboarding, then supervised technology at home, with remote monitoring keeping clinicians in the loop. That architecture is spreading across neurorehabilitation and is the backbone of virtual-first chronic pain treatment, including Karuna's program, where a headset ships to the patient and coaching happens remotely across twelve weeks.

For chronic pain specifically, the maintenance framing translates naturally. Retraining a sensitized pain system, like retraining a hemiparetic gait, is not an event but a practice, and practices survive on convenience. Care that lives where the patient lives, the theme running from bedside MRI to this study, is how practices survive. Our guide to VR pain management covers what the practice consists of.

Frequently asked questions

Did home VR beat usual care in this study?

During the three-month intervention itself, no, between-group changes were not significantly different. The separation appeared at six-month follow-up, where the home-VR group maintained walking capacity while the usual-care group declined significantly.

With eleven people in the home group and a non-randomized-style 1:2 allocation, that durability finding is promising rather than definitive. Its natural interpretation is maintenance: home VR kept gains that otherwise eroded.

Was home VR safe for stroke survivors?

In this study, yes: no study-related adverse events occurred in either the clinic or home phase. Participants were screened, they needed sufficient cognition (MMSE at least 20) and the ability to stand for 30 seconds, so the result speaks to appropriately selected patients using a tele-monitored system, not to unsupervised use by anyone.

How much did participants actually use the system at home?

Adherence was 89%, roughly 55 of 60 planned sessions, with a mean of 35.3 minutes of VR activity daily, divided across motor, cognitive, and occupational exercises. Usability scored a median of 77.5 out of 100 on the System Usability Scale.

Those engagement numbers, from participants with a mean age of 59, are the study's strongest and most transferable finding.

Why did everyone improve in the clinic phase if these were chronic patients?

The clinic phase had no control group, so the cause can't be isolated. Supervised practice, attention, expectation, and the program itself all plausibly contributed. What makes the gains noteworthy is the population: at six-plus months post-stroke, meaningful improvement is not the default trajectory.

It took the controlled home phase, and its follow-up, to produce a comparison, and that comparison is where the maintenance signal emerged.

Does this study support virtual care for chronic pain?

Indirectly. It is stroke research, and its outcomes are motor and cognitive, not pain. What transfers is the delivery evidence: a headset-based program with remote monitoring can sustain months of near-daily home practice safely in a middle-aged-to-older clinical population.

That is the feasibility foundation any home-based retraining program stands on, including VR programs for chronic pain, whose own mechanism and evidence are discussed in VR pain management.

Sources & research.

  1. Jonsdottir J., Baglio F., Gindri P., et al. (HEAD Study Group), Frontiers in Neurology, 2021. Virtual Reality for Motor and Cognitive Rehabilitation From Clinic to Home: A Pilot Feasibility and Efficacy Study for Persons With Chronic Stroke. 12:601131
  2. Trial registration: ClinicalTrials.gov NCT03025126

Related guides.

Free consultation

Ready to unlearn your pain?

Talk with our care team about your pain, your history, and whether KVET™ is right for you. Free, and from the comfort of home.

Free. No symptom or health details are collected on this form.